When coders & care teams collaborate
Three days. 21 people. One heck of a brainstorm.
By: Amber Greene, Kyle Decker, and Neves Rodrigues, Member Experience Research & Product Design at Cityblock Health
Charrette /SHǝ’ret/ noun: A meeting in which all stakeholders in a project attempt to resolve conflicts and map solutions. Requires a heavy supply of caffeine, sticky notes, and enthusiasm.
A three-day, 21-person meeting is many peoples’ worst nightmare. But we like to take risks at Cityblock — so, armed with sticky notes and boundless ideas, we took a deep breath, ordered a vat of coffee (or ten), and hoped the team would be up for the challenge. In true Cityblock style, everyone came together to collaborate on equal footing, whether we were physicians or engineers, Community Health Partners or designers, joining forces to prove that the whole could be more than the sum of its parts. Our mission: reimagine entirely how our care teams manage tasks, collaborate, document, and prioritize work in Commons to best care for our members.
We were all excited by the potential of our first design charrette: to tackle such a thorny challenge that spans the full range of our organization, not with cautious iteration or slow negotiation, but by rapidly establishing shared context, vision, and ideas. We also aspired to find an elegant design solution to this challenge, recognizing that there are always trade-offs and we must make strong bets, rather than falling into design-by-committee.
The challenge
Health care is messy — most of us have experienced that firsthand. And one of the messiest parts is organizing complicated and often disjointed information about members’ health, including not only medical conditions but also behavioral health and social barriers. Even more challenging is translating that multifaceted data into personalized care plans that meet each member’s specific needs. That’s what we’re working to solve every day at Cityblock.
Since inception, Commons, our care delivery platform, was built to enable our care teams — primary care providers, nurse care managers, behavioral health specialists, Community Health Partners, and others — to deliver better care for our members. This is our top priority, and what we’re working towards every day.
One of our recent challenges: how do we make care planning easier for our teams? In other words, how can we make it smoother for our care teams, who are today our primary end users, to map out members’ priorities for their care in the day, month, or even a year ahead? And how do we make our care plans as member-driven, personalized, and impactful as possible?
Before we could tackle that challenge, we first needed to have an accurate representation of all of the moving parts care teams were dealing with — every open task, every goal, and every priority across their panel of members. That set of information, if complete and accurate, could be the foundation for building sophisticated care planning and coordination tools — and reduce a lot of headaches across our care teams.
As we started thinking about this challenge, a flood of questions arose, building on our experience iterating on Commons over the last year: how might we get comprehensive, accurate, and timely information in Commons? How might we reduce the time spent documenting and increase time spent with members? And how might we minimize the friction of maintaining data in Commons, knowing that if care teams find it immediately useful to keep Commons up-to-date, they will do so? We needed to answer those questions to ultimately address our users’ needs.
The process
This was no simple challenge — certainly not one that the Product team could solve alone. True to Cityblock’s co-design roots, we knew we needed to draw on a wide range of expertise at Cityblock: clinical, operational, product, design, research, engineering, and beyond. We would need to establish shared context so people with myriad expertise would deeply understand our core questions. We would also need to build rough consensus on the path forward so that decisions could be lasting and dependable.
For all these reasons, we decided to organize a three-day meeting to bring all key voices together. A shared conversation could establish shared context and allow for a range of contributions — and therefore, allow us to reach working consensus. We drew on the GV Design Sprint guide and our own experiences across a range of organizations to organize the agenda.
Day One: creating shared context
We walked through the planned agenda, and then the group anonymously gave voice to hopes and fears, both for the charrette and the product more generally. One hope expressed the opportunity particularly well: “I hope that Commons will give more than it asks for.” Then the product team spent several hours interviewing clinical and operational participants, learning how they manage tasks, deadlines, and priorities today. Some participants drew illustrations on a whiteboard. Others presented Commons on a laptop, using dummy data to ensure HIPAA compliance.
“I hope that Commons will give more than it asks for.”
Perhaps the most important insight was just how much of our care product work is reactively driven by member emergencies and episodes, rather than rigid plans defined hermetically in advance. Another core insight: our care teams write notes after each member interaction, but those notes are not consistently used to drive proactive care planning. As one participant put it, “In a crisis, I’ve usually completed the task before I’ve even added it.”
The quick take:
Kyle, Product Design
On day one, our care teams shared how they go about their daily work. Many of them had come up with inventive ways to bucket members and tasks in spreadsheets and notebooks — hacks created to get around a product that wasn’t quite working for them. Those hacks are now directly influencing new designs.
Day Two: creative divergence and exploration
Drawing on insights from day one, on the second day we focused on creativity. We used sticky notes to brainstorm a range of possible technology solutions for task and panel management, and collaborated in small groups on storyboard sketches. We ended up with a couple of dozen ideas. Some ideas were very focused on structured data, others bet on free-text narratives; some ideas were focused on audio, yet others on writing or visualizations. Not all were in tension, though — in fact, many could be complementary.
The quick take:
Neves, Product Design
At the start of day one, everything is very fragmented, very elusive. You have the faintest of thoughts converging. As things get going, it all starts coming together. The solutions are creative, a collision of minds and a confluence of perspectives. Given how much we care about our users, creating something that would be of value to them is deeply rewarding. We all know that at the end of the day what we build will end up making our members lives a little better.
Day Three: reasoned convergence and synthesis
We voted to divide near-term priorities from ideas with longer-term potential. Then we used color-coded sticky notes to flesh out pros, cons, and open questions for each promising idea. We voted on those ideas too and built a shared understanding of the basic trade-offs between ideas. One core tension was between managing around our care team’s interactions with members — one of which could be five hours long, with many pieces of insight and many next steps — versus managing around a specific member goal being served (e.g. “John’s diabetes management,” which could permit tighter collaboration, tracking, and resolution). Ultimately, we realized that those two approaches could be complementary parts of a single workflow, and ended the third day with a clear path forward.
The quick take:
Amber, Member Experience Research
Having a productive and inclusive meeting with 20+ people is actually possible! It requires thoughtful planning, intentional facilitation, and some anxious thoughts about how it will all come together, but seeing the inclusive collaboration and high level of engagement made it worth it.
The outcome
In the weeks following the charrette, we created designs to represent our planned approach. Since a picture is worth a thousand words, here are some sketches and how they translated to Commons:
What we learned
As the dust began to settle on the charrette, we sent a feedback survey to participants. A supermajority gave it high marks and encouraged us to use a similar format to tackle future complex challenges. Participants particularly valued the range of stakeholders present — co-founders, Community Health Partners, engineers, and others — all engaging in a fundamentally democratic manner, and the opportunity to build shared context on the first day.
#1: Sticky notes are pretty powerful.
To give every participant a voice, we used sticky notes in unusual ways. For example, to establish shared context on the first day, we wanted the group to swap notes on hopes and fears. Unfortunately, the large group precluded us from having each person speak in turn. So we asked every participant to take a stack of sticky notes and write one hope per note; then we asked people to shuffle the notes in a pile and pull some out at random; finally, we asked people to silently read the ones they’d drawn and pass them clockwise.
Thus, each person read a random sample of other peoples’ hopes. We then opened the floor for anyone to share reactions to patterns they observed in the notes. This method allowed everyone to engage, allowed people to express themselves anonymously, and allowed the group to establish shared vocabulary as well.
#2: Build consensus on the pros and cons of ideas.
To move beyond voting, on the third day, we were attempting to filter and synthesize a few dozen specific ideas. A common design technique would be to give participants small dot stickers and ask them to vote on their favorite ideas. We used a common design technique — giving people dot stickers to vote on the ideas — but then we switched to a more meta conversation.
We asked participants to articulate pros, cons, and open questions for each finalist (using different color stickies to color-code pros, cons, and questions), and then use dot stickers to vote on pros, cons, and questions they agreed with. This process avoided a “lowest common denominator” conversation, and instead supported a more sophisticated analysis of strengths, limitations, and contingencies of various ideas. Ultimately, this enabled us to combine elements of ideas into the final direction.
#3: Diverse experiences translate to great ideas
Though the challenge seemed daunting and the process complex, a three-day charrette was undeniably time well spent. Bringing coders and care teams together, sharing their knowledge and experience — and translating that to a period of shared, focused planning allows us to accelerate progress. And, in turn, it enables us to better serve the needs of our members — which, after all, is why we are here.
Build better care with us (we promise it’s not all three-day meetings!) — we’re hiring across roles! cityblock.com/careers
*All information shown here is dummy data used for testing & learning purposes.
